Considering the early role of human faces as compelling stimuli for typically developing newborns and considering the emergence of sophisticated face recognition skills by six months in typical development, it is very likely that face processing in Autism Spectrum Disorder (ASD) is affected from the first months of life. This project aims to characterize the emergence of face processing abnormalities in infants with ASD by prospectively following, from 3 months to 36 months of life, 160 infants at high risk for developing ASD due to a genetic liability (i.e. by having an older sibling with ASD). At 36 months, clinical diagnosis of the high-risk population will afford the division of the population into three groups: 1) siblings with ASD; 2) Non-Affected siblings (NAF); and 3) siblings with various developmental delays but without ASD features (DD). Developmental changes in face processing skills will be compared against 40 low-risk, typically developing infants (TYP). The set of interrelated face processing skills to be examined include: 1) localization of faces and focused attention to faces; 2) attentional disengagement from faces; and 3) face scanning and recognition. Localization of faces and focused attention towards faces tests the hypothesis that, when presented with faces in naturalistic environments, children who later develop ASD will take longer to localize a face, and, when they do, will spend less time focused on the face. Disengagement from faces task tests the hypothesis that DD, TYP and NAF attend to faces in an obligatory manner and process them on a deeper level as compared to children who develop ASD. This hypothesis will be reflected by a longer latency in shifting attention away from the face to a remote target that appears suddenly. Face scanning and recognition will test the hypothesis that infants with ASD will, with age, exhibit a progressively atypical pattern of face scanning and that the degree of abnormality will correlate with recognition skills as well as social and cognitive deficits. The process of encoding facial features and recognition skills will be gauged via an infant-controlled habituation paradigm. It is expected that the patterns in localization of faces, focused attention towards faces, attention capture by faces, and face scanning and recognition, will, in combination, highlight the developmental divergence in high-risk and low-risk children and correlate with worsening autistic symptoms. Our goals are to identify the type of primary impairments in face processing and to understand the developmental trajectories of face processing abnormalities in the earliest stages of ASD. This will inform early diagnosis (through identification of potential early indicators of risk for ASD), treatment (though identification of pivotal target skills), and theory (through enhancing understanding of mechanisms underlying abnormalities in face perception in ASD). This project addresses several key action items of the NIH Interagency Autism Coordinating Committee, with emphasis on understanding of underlying mechanisms of social impairment and identifying behavioral markers for ASD in infancy.